Production of HXMT calibration Files Outline -...

Outline

25-28 March 2013 IACHEC

Production of HXMT calibration Files

8th IACHEC meeting

LI Xiao-bo Institute of High Energy Physics, CAS 1


Outline

1. Hard X-ray Modulation Telescope(HXMT)

2. Flow of calibration data files

3. Calibration parameters for energy response matrix – Energy scale

– Energy resolution

– Efficiency

4. Monte carlo simulation based on Geant4

5. Performance of HXMT detectors in electronics test

6. Summary

25-28 March 2013 IACHEC 2


1. HXMT

• Scientific Objectives:

X-ray all sky survey

pointing obs. of X-ray source

• Orbit: circle, 550km

orbital period: ~90mins

orbital inclination: 43 degree

• Instruments:

HE: NaI(Tl)/CsI(Na), 5000cm2, 20-250keV;

ME: Si-Pin, 952cm2, 5-30keV;

LE: CCD, 384cm2, 1-15keV

ME HE LE



2. Flow of calibration data files

• detSim

– Inputs:

• Mass model of instruments and environment

• Some commands to configure the simulation

– Outputs:

• Raw event files(ROOT files)

– External dependencies:

• GEANT4—toolkit for MC of particle interactions with matter from CERN

• ROOT– Data analysis package from CERN.


detSim

Mass

model

Comds. root

files specSim

Cal.

Params.

rsp

db

CALDB

rspG

en

Data

Analysis

software

Spectrum,

efficiency


• specSim

– Inputs:

• Raw event root files

• Calibration parameters from experiment data

– Outputs:

• Energy response curve and efficiency

• FITS file, like rmf, arf.


• ROOT– Data analysis package from CERN.

• CCFITS—FITS data file I/O

• rspGen

– Inputs:

• FITS file from CALDB

• Some specific FTOOLS developed by HXMT group to generate the RMF of observe ID.

– Outputs:

• Applicated RMF files


• CCFITS—FITS data file I/O

• FTOOLS of CALDB related.

Continue:


• Energy scale and resolution can be gained through radioactive sources and X beam experiments on ground.

6

3. Calibration parameters

241Am@ HE

• 59.5 keV full energy peak fit: • The mean and sigma of full energy peak are influenced by the fit range and fit

function which we used to describe the background. • We change the different fit conditions ten more times and get the value and error

of full energy peak and its sigma.


• Fit function for the energy-channel relation:

• C=K*E+con, nonlinear etc.

• The error of energy scale and resolution can be derived through the error propagation .

7

C=2.515*E+1.614

NIM A675(2012)78-83

σ/E=0.006+0.47/sqrt(E)


• One method is to use calibrated radioactive sources – The reference activities were provided in a calibration certificate by

the supplier of the radioactive sources.

– The activity at the day of measurements were calculated by taking into account the time elapsed since the calibration reference day.

– N is the detected events, A is the activity of the source, T(E) is the transition probability of entrance window, p(E) is branch ratio.

– For HE, N must meet the energy cut(20—250 keV) and pulse width cut to get NaI-like events. For ME and LE, N should meet the energy cut.

– Error of efficiency can be estimated.

8

Efficiency calibration

𝜀 𝐸 =𝑁

𝐴 ∗ 𝑡 ∗ 𝑝(𝐸) ∗ 𝑇 𝐸 ∗ 𝑒𝑥𝑝 − 𝜇𝑎𝑖𝑟 ∗ 𝑑𝑎𝑖𝑟 ∗ Ω/4π


4. MC simulation based on G4

Incident Generator

Geant4 Mass Model,

Physical process model

Hit information Deposit energy

Deposit position

Digitization ADC spectrum

Pulse width spectrum

Energy resolution and scale

All kinds of incident particles

Exp. environement

Phyical process model

Resolution

curve

Incid-ent

source

Mass model

• Output hit information of MC:

• Parameters for MC tuning:


Visualization of HXMT detectors

SCD detectors of LE in one box:

Collimators of LE: Collimators of ME:

Si-Pin detectors of ME in one box:


MC Spectrum of HE detector

detSim—raw physical spectrum specSim—instrument-like spectrum

Energy scale and

resolution

Pulse width spectrum: HE detector visualization:


• The difference between MC spectrum and data decides the precision of the energy response.

• Experience to tune the MC parameters to reduce the difference . • How big system error will be produced if we have several percent

difference between the MC spectrum and data?

12

MC energy spectrum vs Data

Data MC

241Am 59.5keV @ HE


HE Detector response:

Efficiency:

Energy threshold:>20keV

RMF:


5. Performance of HXMT detectors now

Energy spectrum of 6 detectors in one group. One color represents one hour.


Pulse width spectrum

15

f

𝑓 = (𝑀2 −𝑀1)/(𝜎1+ 𝜎2)

100 hours of test for one group of HE.


Automatic gain control

• 241Am source of in-orbit calibration detector was used on one PMT in the 100 hours test.

mean

sigma


• Time Interval distribution of 6 HE detectors in one group. • Time used to process one signal needs 12us.

17

All Events, including GPS second events, physical events and radioactive events.

Physical events or radioactive events

Time interval distribution


6. Summary


• Elaborate calibration experiments will be done in IHEP soon.

• Detector simulation: – Simulation software – Preliminary MC result vs experiment data – Long-term work to tune MC parameters

• Calibration parameters and their accuracy – Energy scale, resolution and efficiency – Timing related

• CALDB – Design the CALDB of HXMT – Calibration software for users

18

Thank you for your attention!

Production of HXMT calibration Files Outline -...

Documents

Transcript of Production of HXMT calibration Files Outline -...